1. Field of the Invention
The present invention relates to a radiography apparatus.
2. Description of the Related Art
Conventionally there exists a medical X-ray still-image capturing system, where X-rays illuminate a patient, and the transmitted X-ray image is exposed on a film. Films are available worldwide, are light and easy to handle, serve to display and store information, can be made so as to have a larger area, and can have a high gray scale range. However, film requires a developing step, a storage place for a long period of time, and time and work for retrieval.
A conventional moving-image capturing system includes an image intensifier (hereinafter, abbreviated to I.I.) is in the mainstream. The I.I. is generally highly sensitive (e.g., from using photomultipliers) and thus advantageous from the viewpoint of a lower required exposed dosage. However, the I.I. distorts a peripheral image, from a typically attached optical system, has low contrast, and is large in size. With the I.I, not only a fluoroscopic image of a patient is monitored by a medical doctor, but also an analog output of a charge-coupled device (CCD) is digitalized so as to be recorded, displayed, and stored. However, since high gray scale is used for medical diagnosis, even when the I.I. is used for capturing a fluoroscopic image, a film is often used for capturing a still image.
A demand for digitalizing an X-ray image in hospital has recently increasing. In order to satisfy this demand, in place of a film, a radiography apparatus (i.e., a flat panel detector (FPD)) having a structure in which an X-ray detecting element has image capturing elements arranged in a two-dimensional pattern is used and an X-ray dose is converted into an electrical signal. Since this allows an X-ray image to be replaced with digital information, image information can be instantly transmitted remotely. Hence, a patient even residing in a local area can enjoy a high level of diagnosis equivalent to that in a university hospital located in the heart of a city, while the filmless operation saves a film-storing space in a hospital. If an excellent image-processing technique can be introduced, an automatic diagnosis with the aid of a computer and without diagnosis of a medical doctor is possibly.
In recent years, a radiography apparatus (a flat panel detector) including image-capturing elements, each composed of a non-monocrystalline thin-film semiconductor such as an amorphous-silicon and capable of capturing a still image has been commercialized. With the technique for preparing a non-monocrystalline thin-film semiconductor, the detector having an area greater than 40 cm squares, covering the size of a human breast has been achieved. Also, since its preparing process is relatively easy, offering the detector at low price is expected. Besides, since a non-monocrystalline thin-film semiconductor such as an amorphous-silicon can be prepared in a form of a thin sheet of glass having a thickness not greater than 1 mm, the detector has a very thin profile.
Such an X-ray radiography apparatus includes a conversion circuit having a plurality of conversion elements arranged therein in a matrix pattern, each configured to convert X-rays into an electrical signal, and a reading circuit configured to read the electrical signal from the conversion circuit.
FIG. 10 illustrates the two-dimensional circuit configuration of a known conversion apparatus. The conversion apparatus includes a photoelectric conversion circuit unit 701 and a reading circuit 702. The photoelectric conversion circuit unit 701 includes photoelectric conversion elements S1-1 to S3-3 serving as conversion elements, switching elements (thin film transistors (TFTs)) T1-1 to T3-3, gate wiring lines G1 to G3 configured to turn on-off the TFTs, signal wiring lines M1 to M3, and a wiring line Vs biased by a power supply Vs and configured to apply an accumulation bias on each of the photoelectric conversion elements. Also, the conversion apparatus includes a shift register SR1 configured to apply a driving pulse voltage on each of the gate wiring lines G1 to G3. A voltage Vg for turning on-off the TFTs is externally supplied.
The reading circuit 702 amplifies outputs of parallel signals of the signal wiring lines M1 to M3 in the conversion circuit unit, converts them into a serial signal, and outputs the serial signal. The reading circuit 702 includes switches RES1 to RES3 configured to reset the signal wiring lines M1 to M3, amplifiers A1 to A3 configured to amplify the signals of the signal wiring lines M1 to M3, sample-and-hold capacitors CL1 to CL3 configured to temporarily store the signals amplified by the amplifiers A1 to A3, switches 5n1 to Sn3 configured to perform sample-and-hold, buffer amplifiers B1 to B3, switches Sr1 to Sr3 configured to convert parallel signals into a serial signal, a shift register SR2 configured to apply a pulse on the switches Sr1 to Sr3 so as to achieve serial conversion, and a buffer amplifier 104 configured to output the serial signal.
An operation of the conversion apparatus shown in FIG. 10 will now be described. FIG. 11 is a timing chart illustrating the operation of a conventional conversion apparatus.
A conversion period (e.g., an X-ray illuminating period) will be described. In a turned-off state of all TFTs, when a light source (e.g., an X-ray source) is turned on in a pulsating way, with a wavelength converter (not shown), radiations are converted into light having wavelengths in a wavelength range allowing the photoelectric conversion elements to be sensitive to the light. The light illuminates each of the photoelectric conversion elements, and signal charges corresponding to the quantity of the light are accumulated in respective element capacitors. If the conversion elements are sensitive to the particular radiation source (e.g., X-rays), the wavelength converter can be removed, and the signal charges corresponding to the dose of the radiation source can be accumulated by the conversion elements. Even after turning off the light source, the signal charges subjected to photoelectric conversion are held in the element capacitors.
A reading period will be described. A reading operation is sequentially performed in order from the photoelectric conversion elements S1-1 to S1-3 in the first line, S2-1 to S2-3 in the second line, and to S3-1 to S3-3 in the third line. First, in order to read the photoelectric conversion elements S1-1 to S1-3 in the first line, the shift register SR1 applies a gate pulse on the gate wiring line G1 of the switching elements (TFTs) T1-1 to T1-3. With this operation, the switching elements T1-1 to T1-3 are turned on, and the signal charges accumulated in the photoelectric conversion elements S1-1 to S1-3 are transferred to the signal wiring lines M1 to M3. Since the signal wiring lines M1 to M3 have reading capacitors CM1 to CM 3 added thereto, the signal charges are transferred to the reading capacitors CM1 to CM3 via the corresponding TFTs. For example, the reading capacitor CM1 added to the signal wiring line M1 has a capacitance equal to the total sum (corresponding to three capacitances) of capacitances (Cgs) between gate and sources electrodes of the TFTs T1-1 to T3-1 connected to the signal wiring line M1. The signal charges transferred to the signal wiring lines M1 to M3 are amplified by the amplifiers A1 to A3 respectively. The amplified signals are transferred to and held in the capacitors CL1 to CL3 while turning off an SMPL signal.
Subsequently, when the shift register SR2 applies a pulse on the switches Sr1, Sr2, and Sr3 in that order, the signals held in the capacitors CL1 to CL3 are outputted from the buffer amplifier 104 in order from the capacitors CL1, CL2, and CL3. Since analogue signal outputs of the buffer amplifiers B1, B2, and B3 are outputted from the buffer amplifier 104, the shift register SR2 and the switches Sr1 to Sr3 are collectively called an analogue multiplexer. As a result, photoelectric conversion signals of the photoelectric conversion elements S1-1, S1-2, and S1-3 corresponding to one line are sequentially outputted from the multiplexer. Reading operations of the photoelectric conversion elements S2-1 to S2-3 in the second line and S3-1 to S3-3 in the third line are likewise performed.
When signals of the reading capacitors CM1 to CM3 are sample-held in the capacitors CL1 to CL3 with the SMPL signal for the first line, the reading capacitors CM1 to CM3 are reset at a GND potential with a CRES, and a gate pulse of the gate wiring line G2 is then applied. In other words, during a serial conversion of signals in the first line with the shift register SR2, signal charges of the photoelectric conversion elements S2-1 to S2-3 in the second line are concurrently transferred by the shift register SR1.
With the above-described operation, signal charges of all photoelectric conversion elements in the first to third lines can be outputted. In the foregoing operation of the photoelectric conversion circuit, an X-ray image can be read, however, the actually read image includes an offset generated in the photoelectric conversion circuit and the reading circuit.
The offset is generated due mainly to the following two factors: (A) a dark current of each of the conversion elements, and (B) an offset voltage of each of the amplifiers (e.g., A1 to A3) of the reading circuit. Since an X-ray illuminated image includes an offset component, the offset component can be removed. This removing operation is called an offset correction.
In the case of capturing a still image, the offset correction is performed such that a single sheet of an X-ray-illuminated image is first captured, then, a single sheet of an offset image having no X-rays illuminated thereon is captured, and the offset image is taken out from the X-ray image. Offset imaging is conducted in the same way as X-ray radiographing (i.e., X-ray imaging), except instead of multiple X-ray images one of the images is due to illumination without X-rays. In other words, offset components for all of the pixels must be read to obtain an offset for a single sheet. The foregoing technique is disclosed in U.S. Pat. No. 6,333,963.
In the case of capturing a moving image, the following two methods are offered. According to one method (a continuous imaging method), a sheet of offset image without illumination of X-rays is first captured, then, X-ray capturing is continuously performed, and the X-ray images are corrected with the previously captured offset images. According to the other method (an intermittent imaging method), an X-ray image and an offset image are alternately captured, and the X-ray image is corrected by taking out the offset image therefrom on a basis of its capturing operation. The former presents a high frame rate since only a single sheet of an offset image is captured and, thereafter, allows X-ray images to be continuously captured. However, the latter method has a low frame rate (half the high frame rate) since an-X-ray image and an offset image are alternately captured.
Unfortunately, the case of capturing a moving image currently has the additional feature of an offset fluctuating over time. The fluctuation is discussed in Japanese Patent Laid-Open No. 2002-301053. Here, this feature will be described. According to the description of the foregoing patent document, upon capturing a moving image, especially upon capturing a fluoroscopic image, an offset varies every image capturing operation, causing deterioration in image quality. The foregoing patent document discussed the use of an intermittent imaging method as a countermeasure against the fluctuation of an offset, in which X-ray imaging and offset imaging operations are alternately performed so as to update the offset image.
FIGS. 12A to 12D illustrate a timing chart of a control method discussed in the foregoing patent document, wherein the time axis extends horizontally. The contents of the control method will be briefly described below. First, with “FPD collection” (i.e., collection of X-ray image data), X-rays illuminate an object and are captured in an X-ray image. With “collection of calibration data” (i.e., “offset imaging”), offset data is captured, added to previously collected offset data (not shown), the summed offset data is averaged, and the averaged offset data is updated as new offset data. Then, with repeated “FPD collection”, an image of the X-ray illuminated object is captured and undergoes an offset correction by the updated offset data. As described above, by alternately performing “FPD collection” and “collection of calibration data”, offset data is updated as needed so as to inhibit fluctuation of an offset.
While inhibiting fluctuation of an offset as described above, the intermittent imaging method has a low frame rate.
As a method for increasing the frame rate, a pixel addition method is offered. FIGS. 13A and 13B are timing charts illustrating the contents of the pixel addition method, where FIG. 13A illustrates a normal imaging method in which the three gate wiring lines G1 to G3 shown in FIG. 10 are increased to six wiring lines G1 to G6, and FIG. 13B illustrates an example operation of the pixel addition. The pixel addition refers to a method of concurrently reading signals in a plurality of lines. In FIG. 13B, two gate wiring lines are concurrently turned on, and signals corresponding to the two lines are concurrently outputted. With this method, a reading time is reduced by half and a frame rate is increased by double. However, concurrently outputting signals corresponding to the two lines makes an area of a single pixel double, causing a reduced resolution. While the pixel addition is performed by concurrently turning on the two gate lines in the example shown in FIGS. 13A and 13B, the frame rate can be made triple or quadruple by increasing the number of the gate lines to be turned on to three or four.
As described above, the continuous imaging method has a deterioration in image quality while having a high frame rate, and the intermittent imaging method has a low frame rate while inhibiting fluctuation of an offset. In particular, when a non-single crystal semiconductor, such as amorphous silicon, is used in switching elements, the frame rate is considerably reduced since the transfer time of the switching elements is long. In addition, the pixel-addition method has a low resolution while having a high frame rate.